An emerging machine learning strategy for electrochemical sensor and supercapacitor using carbonized metal–organic framework

Abstract

• A machine learning strategy was applied for electrochemical sensor and supercapacitor. • The carbonized metal–organic framework showed excellent electrochemical response for niclosamide (NA) and good capacitance. • Both adsorption and binding energy were theoretically calculated and the structure of prepared material was optimized. Machine learning (ML) plays an important role in the electrochemical application of electrode materials. In this work, an emerging machine learning strategy for both electrochemical sensor and supercapacitor using carbonized metal–organic framework (C-ZIF-67) is proposed. The morphology and element analysis of C-ZIF-67 are characterized and further demonstrate the presence of C, N, O, Co elements. The ML model based on artificial neural network (ANN) algorithm as a powerful tool to realize intelligent analysis of niclosamide (NA), the derivative technique as an auxiliary means of voltammogram treatment to reduce personal error from data-reading and improve the sensitivity of electrochemical responses at very low concentrations, and the theoretical calculation is employed for both adsorption and binding energy, optimized structure of the prepared sensing material. The developed sensor displays excellent electrochemical response about 196.6-fold improvement compared with the bare GCE for NA, wider linear ranges of intelligent analysis from 1 nM to 9 μM with low limit of detection of 0.3 nM, and satisfactory practicability. ML model with ANN algorithm is also employed for predicting the performance of supercapacitor. The supercapacitor shows good performance with capacitance of 336.67 F/g at the current density of 2 A/g and excellent prediction with acceptable errors. This work will provide a new strategy for the development and electrochemical application of bifunctional electrode materials using the ML technique combined with theoretical calculation.